Long-term performance of epoxy-based coatings: Hydrothermal exposure

We examined the effects of prolonged exposure to high temperature and water on epoxy-based powder barrier coatings applied to steel panels, which are commonly used in many industrial applications including oil & gas pipelines. The coatings’ performance was evaluated over 85 weeks at 65°C using deionized water. We also compared the mass transport properties of free-standing coating films with the barrier performance of the coated steel panels. This research lays the groundwork for predicting cumulative damage and time-dependent barrier performance of defect-free coating systems. Despite the fact that these coating systems are intended for decades of in-service use, we found that degradation effects caused by permeant sorption within the coatings can be detected as early as 8 weeks in the ageing process. The first 200 days of exposure emerged as critical for underlying corrosion reactions, marking the completion of epoxy network degradation and the onset of a steady state in mass transport mechanisms. Despite the protective barrier coatings, we observed readily occurring under-coating oxidation of the steel substrate after 182 days, as confirmed by cross-sectional and focused ion beam milling analysis. We also analyzed the adhesion strength of the coated panels over time. The epoxy-based coating’s pull-off strength declines rapidly due to water-induced plasticization, but the adhesion properties of the epoxy network show a slight recovery due to secondary cross-linking by Type II bound water. This study underscores the complexity associated with predicting the time-to-failure for epoxy coating systems. However, the data and analyses provided herein offer valuable insights into the implications of extreme exposure conditions, aiding in the construction of lifetime predictions using a stochastic process. In real-world scenarios, pipelines undergo various fluctuations in parameters like temperature and humidity, potentially leading to failure. A deterministic physical/chemical model under simplified conditions can serve as input for the probability distribution function of future failure events.